work-with-jev
Local-first work-message classifier that uses Jev to sort chat messages into urgent, to-do, worth reading and skip, with cross-chat to-dos.
Overview
Work with Jev (工作消息分类器) is a local message classifier for busy work group chats, documented in Chinese with an English README. You tell it who you are and what you are responsible for, and it uses Jev to judge whether each message concerns you, needs action, is urgent or is worth reading, sorting messages into four columns: urgent, my to-dos, worth a look, and skip for now. Important items can be added to a cross-chat to-do queue, reordered and checked off.
It connects to Feishu (selected group messages the user authorizes) and WeCom (bot private chats and @mentions in groups), and is designed so new sources such as DingTalk or Slack can be added. A demo with labeled preset results runs without a key. Classification can be corrected manually, original text is kept, and it never replies automatically. The README notes that probabilities are not accuracy guarantees and that thresholds should be checked on your own messages. MIT licensed with offline regression tests.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Choice | Which column does this message belong in for this user? | urgent · my to-do · worth a look · skip for now |
| Yes / no | Does this message require action from me? | yes · no |
How it works
- Set your role and responsibilities.
- Pull messages from Feishu, WeCom or the demo set.
- Jev judges relevance, action, urgency and value.
- Sort into four columns; add items to the cross-chat to-do list.
Strengths
- Personalized by your stated role.
- Separates reading messages from calling the model.
- Demo works without a key.
Caveats
- Default thresholds need checking on your own data.
- Connectors depend on platform permissions.
Getting started
git clone https://github.com/Adkid-Zephyr/work-with-jev.git && npm ci && npm start, then open http://127.0.0.1:4173At a glance
- Type
- Project
- Status
- beta
- Language
- JavaScript
- Runtime
- local Node server + web UI
- Has code
- yes
- Has tests
- yes
- License
- MIT
Use cases and integrations
How curated lists describe it
Work with Jev is a local-first message classifier that uses Jev to sort work messages into urgent, to-do, worth-reading, and skippable groups with cross-chat to-do management and Feishu and WeCom adapters.
logicrw/awesome-jev-projects
Frequently asked questions
What does Jev decide?
Whether each message is relevant, needs action, is urgent or is worth reading for you.
Which chat platforms are supported?
Feishu and WeCom, with an extension guide for others.
Does it reply for me?
No, it only classifies and organizes.
Related projects
Android and Windows message-priority assistant where Jev scores urgency and reply-now for each message and sorts them by colour.
ProjectbetaKotlinOpen-source AI email assistant; listed as using Jev as an optional classifier for email intent and triage.
ProjectproductionTypeScriptTriage for DingTalk mail and chats where Jev sorts messages into needs-reply, updates, promo, sales or spam with attention flags.
ProjectbetaPythonGoogle Apps Script that labels, archives or trashes Gmail conversations based on Jev answers to your plain-English yes/no questions.
ProjectbetaJavaScriptGmail classifier running on Google Apps Script where Jev decides action-needed, importance and category for each thread.
ProjectbetaTypeScriptWindows version of XianHui that OCRs the visible WeChat window locally and uses Jev to rank which messages must be answered first.
ProjectbetaPython